Two weeks ago ICM produced an unexpected result – they did parallel phone and online polls for the Guardian, and found a five point lead for LEAVE in both online and by phone. A five point lead for leave in an online poll was slightly unusual, but not completely out of line with the then average position of neck-and-neck. The five point lead in the telephone poll was remarkable, given most other phone polls (a) were showing a consistently different pattern to online polls and (b) were showing a robust lead for REMAIN. The poll got a lot of attention, but we reserved judgement a little – it was only one poll, and it was conducted over a bank holiday weekend so perhaps the sample could have been affected.

Today ICM have repeated that experiment and confirmed their earlier findings – their online and telephone polls are painting the same picture, and both have LEAVE with a clear lead. Topline figures by telephone are REMAIN 45%(+3), LEAVE 50%(+5), Don’t know 5% – equating to a six point leave lead of LEAVE 53%, REMAIN 47%. Topline figures online are REMAIN 44%(nc), LEAVE 49%(+2), Don’t know 7%(-2) – also equating to a six point lead for leave. Full tabs for both polls are here.

Following MORI’s methodology switch last week, Martin’s commentary on the ICM website also includes a note about educational weighting. In ICM’s case he says their last two phone polls did have a tendency to have too many people with qualifications… but when they tested correcting for it reduced the weighting efficiency but didn’t actually make any difference to the headline results.

63 Responses to “ICM online AND phone polls show six point LEAVE lead”

1 2
  1. Interesting tactics from Osbourne this evening just when you thought Remain actually might have a plan b (small b) Osbourne has said that disability benefits may be cut if we leave – clearly he has learnt nothing.

  2. So YouGov confirms what ICM was showing.

    I wonder what caused the shift in sentiment? The debate a few days ago? Gisela Stuart and Andrea Leadsom came across more measured than the three Remain women.

    Or could it have been the heavy-handed Schauble comments which came across like a threat?

  3. This is not what people expected. As the campaign gets to the crunch time, the status quo was meant to pick up support. I suspect remain is suffering from two critical problems;

    1) Their own dreadful campaign. Over hyped doom and a focus on the dry numbers, most of which aren’t believed. AW’s point about the polls suggesting people don’t believe them anymore is telling, and is at the heart of why they are falling. Barely a single mention of how the EU helps real people, even as an end to mobile roaming charges kicked in.

    2) Remain has no answer to the immigration issue. Boris called out Cameron (and his own GE manifesto) for the ‘few tens of thousands’ deceit, and Cameron has no answer. Remain is paying the price for the broken promises of the last 6 years on this issue.

    I also think that EU leaders may now wish they had taken the renegotiations a little more seriously – they too have disastrously misread UK opinion. The one thing that was forcibly removed from the table instantly was any notion of change to the free movement of people. This now looks to be a huge mistake that may well take the UK out of the EU.

  4. N. Ireland has done well out of the EU grant wise. To not include them in the sample does seem a bit daft. Obviously, contacting overseas voters wouldn’t be easy so I can forgive the polling orgs that one. Why do you suspect overseas voters would support Remain Dunham?

  5. “New constitutional and political reforms are now pushed forward and 27 states more tightly integrate.

    After 25 years of bullying the UK the market of 450m users eventually forces the UK to join back into the EU ‘on their terms’ and accept without any optouts the euro.


    Yes – very fanciful. There are popular insurrections against integration in many EU countries, and the direction of travel is far more likely to be towards loosening.

  6. Overseas voters – no polls include them (it’s practically impossible to include them). However, on the last electoral register there were only 26,000 of them, so their impact will be negligible (of course there could be a massive surge of overseas registrations… but we’ll see).

  7. Thanks Anthony. But why no N.Ireland sample portion? It’s not like their voters are voting for predominantly sectarian parties, this is the same choice across the UK as a whole. Any thoughts please?


    Interesting read. As an econs/stats man, I’d like to see that model and the data that goes with it!

  9. @ Anthony Wells

    Thank you – on the overseas voters. I would have thought there would have been a large increase in registration for this referendum. So, it is negligible.

  10. DSL
    That was from January. Still very interesting. I wonder if his view is still the same?

  11. DSL
    Quote from that article
    “Low turnouts are more influenced by people who are persuaded by the elites and are part of the political scene. Once the turnout goes up then you have more people who only think about politics passingly and they are more likely to reject the position by the government and other political elites.”

    I’m not sure this is true this time, as most commentators agree that UKIP supporters are the most committed, and they are anti-elite almost by definition.

  12. “But what do most people think? I would guess most people don’t even notice the warnings…”


    Well they scare me, they scare me into not voting. But I wasn’t going to vote anyway…

  13. When’s the next leaders approval rating published?

    I expect that both Cameron & Corbyn will have signifcantly slipped in peoples estimation.

1 2